1. Field of the Invention
The present invention relates generally to scanners and scanning methods, and more particularly to a method for dynamically compensating for a faulty pixel in a scan line of a scanner.
2. Description of the Related Art
An imaging system may include a scanner that is used to generate a digital representation of a document being scanned. Such a document may include any of a variety of media types, such as paper, card stock, etc., and may be regular (e.g., rectangular) or irregular in shape, and on which there is formed, for example, text, graphics or a picture, e.g., a photo, or a combination thereof. Typically, a color scanner generates three channels of data, e.g., red, green and blue (RGB), in generating a scanned image of the document.
When scanning an image using a typical scanner, a uniform response of the image sensor across a scan line is critical to the quality of the scanned output image. Certain aspects of the scanner, such as non-uniform illumination and variable response of each individual sensor element (i.e., pixel), make it necessary to correct each individual pixel's light and dark response. This is traditionally accomplished by applying a gain factor to each pixel to stretch the true white response of the scanner to a nominal value and an offset value to set the true black response of the scanner to a nominal value. Such correction is often referred to as pixel-to-pixel calibration and is utilized to normalize the white/black response of the entire scan line. This operation assumes that each individual sensor element is operational and responds appropriately to different intensities of light. However, a very weak or dead sensor element in the scanner cannot be compensated for by traditional methods. For example, FIG. 1 illustrates a scanner image 10 having a defective streak 12 in scan direction 14 as a result of a defective pixel in the scanner. If such a defect is present on the manufacturing line, the scanner is scrapped, thus lowering the overall yield of the product and affecting cost. If such a defect occurs during the useful life of the scanner, the quality of the scanned output is greatly degraded.
One type of faulty pixel is a corrupt pixel. A corrupt pixel, e.g., a very weak or dead sensor element, in the scanner may decrease the performance of an imaging system that utilizes compression on the pixel-to-pixel calibration data, since the white/black response for the pixel is very different than that of the local neighbors of the pixel. Outlier sensor elements may drastically reduce the capabilities to compress the stored calibration data, which has a detrimental effect on the performance/throughput of a device that is bandwidth starved.
Another type of faulty pixel in a scanner is a missing pixel. For example, Contact Image Sensors (CIS) scanners are typically manufactured by stacking multiple sensor chips together in-line. If the boundaries of these chips create a dead space equal to the size of one sensor element (pixel), then the image data in this area is not captured by the scanner. This is essentially a missing pixel.
What is needed in the art is a method that automatically detects a faulty pixel and compensates for the faulty pixel to reduce or eliminate visible artifacts that otherwise would be present in the output image.